Normalize Don’t Improvise:  Drainage, Drainage, Drainage
Data Science & Analytics

Normalize Don’t Improvise: Drainage, Drainage, Drainage

Rosemary Jackson  •  

Petro.ai: Enhances traditional type curve, lateral length and EUR analysis with normalization for drainage

Petro.ai: Unconventional thinking for an industry stuck in the conventional

You’ve got your Pad Complex, or Cube, a 3 dimensional, multi-layer tapping of the oil and gas in shale. Tough, low permeability shale. Now, you’ve got to put one more well, no two, well maybe even 5 into this already pin-pricked earth. To make that decision do you turn to type curves, EUR, lateral length and/or maybe proppant? Those are key variables. But there’s something more important. Petro.ai knows that drainage area is what matters. Well spacing is noisy. Normalize for the drainage area. 

The industry loves a good type curve. That’s because we have that ‘conventional’ frame of mind. Type Well Profile (TWPs) also called type curves apply mathematical adjustments to data like lateral length, normalized first month production. Then scaled by formation or operator, they have become the predictive production tool for the economics of the cube. Until Petro.ai.

“Years ago when we were putting in single wells,” Troy Ruths, CEO of Petro.ai explains, “type curves did a pretty good job because we weren’t condensing the spacing. But now that the spacing is so dense, there’s not enough information in that type curve to know where you’re actually draining. You can’t keep compressing all these wells down into a tighter and tighter cube and expect the productivity to stay the same. The type curve doesn’t differentiate the information finely enough at this point in the production.”

“Engineers are overestimating the resource in place because they do a type curve times well spacing and they’re not taking into account the fact that they’re draining a different area,” Ruths continues.

“They don’t actually know if it’s representative of how much resource they have in place. And if anything, we’ve seen that they’ve been very wrong.”

And the error in estimation doesn’t stop there. According to Kyle LaMotta, VP of Analytics, “We have to stop thinking about EUR. This is the other reason type curves are misleading. Rather than EURs, think about 12 month and 18 month production. If we’ve seen anything, people cannot predict years into the future and modified Arps doesn’t do a great job of capturing the complexity of shale development. You’re better off trying to get your money back in 12 or 18 months. The best companies do that.”

Petro.ai puts together a decision model based on drainage volume and a complex set of geomechanical steps including microseismic, SHmin and SHmax, and the elusive well spacing. Ruths continues, “When you put these wells in and you stimulate, we’re putting them so tightly now that these wells are stimulating each other and there’s a lot of cross talk within this cube.

“As you make these wells come closer, you’re tightening the well spacing. Now, where I think well spacing is incorrect, is we’re overly simplifying a complex problem. We’re oversimplifying because we’re using one variable like the horizontal spacing when in actuality this is a 3D reservoir and these wells are porpoising in the zone. Talking about well spacing is really a proxy for drainage. Well spacing is this massive proxy that the industry has created for drainage.  

“And what they’ve done is tie type curves to well spacing. Now, in effect you’re saying this type curve is draining this amount of reservoir. But because you’re using well spacing, you’re not able to get that math out exactly. It’s a very crude way to measure how much you’re actually draining from the reservoir. And we know that these fracs are propagating out of the zone. And the propagation of the fracs isn’t really due to the intensity of the completion but rather what the geology is giving you. That’s the stage for well spacing as I see it.

When people talk about parent child, when they talk about well spacing, they’re really talking about drainage. They’re talking about drainage and they’re using different terms to talk about that drainage. At Petro.ai what we’ve done is try to get closer to calculating drainage. And recently, with our Permian company clients, we’ve shown that it all adds up to the drainage of that reservoir if you’re a parent or a child well.

“Lateral length right now and well spacing are the proxy for drainage volume. So if you think about a rectangular prism, you have the length of the rectangular prism like 12,000 feet. That’s lateral length. And then you have well spacing which is sort of like the width and the height. That’s the spacing they’re assuming when they space their wells. What Petro.ai is doing is adding a layer of sophistication that actually calculates that area better. We can use that as a normalization variable not lateral length. Instead we use the microseismic approach and understanding frac geometries.

PETRO
114 Main Street Ste. 200, Houston, TX 77002
4-5609 Avenue du Parc, Montreal QC, H2V 4S8

© 2020 Petro.ai All Rights Reserved